Advertisement
The Problem With Personality Tests
Get articles like this
in your inbox
The longest running and most trusted source of information serving talent acquisition professionals.
Advertisement
Related Articles
It seems counterintuitive. Some people claim that personality has everything to do with job performance. Others know that every piece of serious research shows personality scores have almost nothing to do with job skills. I’ll repeat that: personality scores seldom, if ever, equal skills.
Soooo, if personality has nothing to do with skills, why do so many organizations used it to predict job performance? The answer depends on how and where you look.
Let’s start at the beginning. Researchers agree that adding a decent job-related developed personality test will improve hiring success only a teensy-weensy bit … but, only if studies show it directly affects job performance. Aside from that? Nada!
Personality scores and KSAs are totally different things. I’ve done quite a bit of personality research and can assure you that personality scores and personal skills are two different things. In fact, most all serious investigations shows that there is almost no relationship between scores on a personality test and KSAs (e.g., teamwork, conflict resolution, intelligence, learning ability, analyzing, planning skills, and so forth).
So now we have two pieces of trustworthy data: 1) personality scores have almost nothing to do with predicting job performance; and, 2) scores have almost no relationship with KSAs.
One reason is semantics. Personality scores and skills often have the same name: i.e., analysis, teamwork, drive, and so forth. This makes it easy to confuse self-descriptions with skills. Another, as mentioned earlier, is the false belief that a self-description correlates with having a skill. And a third is caused by examining the wrong population.
You see, when people who are already working in a job all have enough skills to stay employed, small differences like personality tend to stand out; that is, job-group members are successfully employed and doing the same kind of work (i.e., a restricted-range population). Real proof, however, is obtained when you give everyone a personality test before hire, ignore the scores, and compare scores to performance six to nine months later. Unfortunately, all the vendors I know use restricted range populations and tell their clients their test predicts job success. You can decide for yourself whether they are being intentionally deceitful or just don’t know any better.
If personality is not a good hiring predictor, what is? Tests, exercises, and simulations where the applicant has to demonstrate (not tell you about) job skills like being smart enough for the job, having the right interpersonal skills, having the right technical knowledge, and so forth.
Can you trust testimonials? ‘Fraid not. Smoke-and-mirrors vendors plaster testimonials all over the place. I think they believe their own hype. Always remember a free opinion is worth every penny you paid. Often their claims are enough to make one weep with either laughter or sorrow. Laughter, if you believe in claims like “90 percent effective,” “fake proof,” “totally reliable,” and so forth. Sorrow, if you count all the qualified people who were rejected and all the wrong people hired based on bogus test scores.
You might notice I always put test scores before job performance, and not the other way round. Clustering employees into high-low groups, giving them a personality test, and averaging scores is akin to throwing stuff on the wall and seeing what sticks. That’s because employees in the same job-group are often highly similar; performance ratings are not as simple as being “high-low”; group classification has more to do with opinions than facts; individuals in a group may not match their own group average; and, there is still no assurance personality caused performance. If your current vendor used a high-low approach, Step 1 is to examine how well individuals in each high-low group actually matched their group average, and Step 2 is to find a new vendor because your old one doesn’t know what he/she is doing … better yet, start with Step 2.
Does anyone really think applicants are totally honest? I know some vendors who claim their tests contain lie scales that identify fake responses, But as a professional test developer, I can say unequivocally that the only absolutely fake-proof test is one with no questions. No one can never know whether a test-taker is making him/herself look good, second-guessing a job-profile, or delusional.
People often get “wowed” by scores in a workshop and assume it will work as a hiring tool. But how much wow magic does it take to add scores from five questions about being outgoing and social, and generating a report stating someone is “extraverted”? Still wowed? Take a minute to Google the “Forer Effect” and read how people almost always agree with their personality reports. As a side note, I was once invited to take a vendor’s test for a spin. To test its accuracy, I chose answers randomly. The vendor proudly claimed the profile report fit me perfectly! I felt bad so I offered to help make the test better. It refused. Good marketer. Bad test developer.
I’ll be the first to admit there are personality conflicts at work. But with whom? With today’s manager? With tomorrow ’s manager? With the work group? With the organization? Personality can certainly affect performance ratings once you are on the job, but it won’t predict skills beforehand. They are, however, very good at producing clones — and cloning has a long history of contributing to organizational disasters.
Some users claim they use personality scores for information only, and not for decision making. Sure. Did we all just fall off the turnip truck? If test scores are information only, give the test afterhire, not before.
I can guarantee with 100 percent assurance that anyone who intelligently tracks hiring results will soon discover which test is bogus and which one actually predicts job performance … then they can add up the wasted dollars spent on employing poor performers.
Personality tests, in even the best cases, are better left in training workshops where they can help people understand differences. Using them to make hiring decisions will always lead to turning away skilled candidates and hiring unskilled ones.